Answers are grounded in this paper and cite their sources.
The relevance . Generative artificial intelligence (AI) is changing the usual parameters of determining the truth. It generates text, images, sound, and video that are close to imitating traditional forms of evidence. However, scientists who access these data often lack confidence in their origin and lack understanding in ensuring their reliability. The issue of responsibility for the materials created by these tools is extremely acute. The purpose is to examine how the results obtained by AI are changing the nature of truth, authenticity, evidence, and trust in modern digital environments. Objective : to conduct a critical analysis of the impact of generative AI on documentary validation and academic integrity, as well as on evidence and institutional responsibility in conducting research and using materials in audiovisual form. Methodology . The research is based on an interdisciplinary analysis of philosophical, sociological, and legal literature. It also takes into account the auxiliary corpus of modern publications in the field of higher education, academic ethics, pedagogical applications of AI and deepfakes. Results . The analysis shows that generative systems do not eliminate the truth, but only destabilize its everyday indicators. The meaning and coherence of the text, the realism of the image, or the naturalness of the voice are no longer sufficient signs of authenticity. The Aristotelian definition of truth as knowledge corresponding to reality does not relate to generated content. Video and audio generators construct plausible, time-consistent sequences. The formal document here is a simulation of competence, presence, or testimony in all these cases. The danger lies in the persuasiveness of the presentation. False information presented authoritatively can be disseminated as knowledge without any justification. Critical competence requires not only doubt about the tool, but also methodical verification of its results. Conclusions . The formulation of evidence for final statements in scientific research should be an open process involving human verification, technical traceability, institutional rules, critical professional training, and legal responsibility. This approach is proposed to be called the "verification mode".
No takes yet. Share an insight, caveat, or question.
Douyon et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: